| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
|
|
|
|
|
| |
Also corrects the 5.0 backport of
1bd1c4f6a0d571443a80c52d1b3f284a0c078af4 from SERVER-59812 and leaves
the ReshardingMetrics intact until the resharding data replication
components have quiesced.
(cherry picked from commit 34cac37ac5a61946aae9d149c8cb2f1d109e7320)
|
|
|
|
|
|
| |
all members in replSetGetStatus
(cherry picked from commit 0cceba6f04b95f3652de84c2f1f4ab2a644dba6e)
|
|
|
|
|
| |
(cherry picked from commit e2089fc73a76db6bcf2ca2729e7e886258d3ad2f)
(cherry picked from commit c80ff81b19a3c5d28892c02f380e8c12ad0a0253)
|
| |
|
|
|
|
|
|
| |
batches on secondaries
(cherry picked from commit 4a230c070cb187604d07d04598a912b4feca937d)
|
|
|
|
| |
(cherry picked from commit d9fcd9f124ece9ab0b3a3c46cb6d7052b7282dd2)
|
|
|
|
| |
(cherry picked from commit 2d3592d80fb851e00b519bfeeea513aaff6d0c95)
|
|
|
|
|
|
|
| |
ReshardingMetrics::onCompletion() can be called multiple times within
the resharding::WithAutomaticRetry blocks of the ReshardingCoordinator.
(cherry picked from commit 5d18bd88c941964e19622282cd040eadbb0db23d)
|
|
|
|
|
|
| |
collection path
(cherry picked from commit bdc2e9b2ed299f4dfbf6183eed94707afbde8478)
|
|
|
|
|
|
| |
handle concurrent priority takeover.
(cherry picked from commit d2d7164bb02fff924ac787f6d422dffb22ef0b94)
|
|
|
|
| |
(cherry picked from commit 651f4a692aa59353954d3cd18ec3572fa9bf48aa)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
ShardsvrCommitReshardCollectionCommand's commit() call and
ShardsvrAbortReshardCollectionCommand's abort() call each perform a
write which will trigger the state document to be deleted on the donor
and recipient. To verify this is done, those commands perform a read to
check if the state documents still exist. Now that the RSTL is not
acquired during reads, it is not guaranteed that the command's opCtx
will be interrupted when performing a read despite calling
setAlwaysInterruptAtStepDownOrUp(). As a consequence, it is possible for
the command's write to have been interrupted due to a step down on the
donor/recipient, causing the document to still exist during the
verification read and triggering a uassert. To resolve this issue, the
commands now do a no-op write before the verification read, ensuring
that the first write has indeed completed without being interrupted.
(cherry picked from commit cca75006b85690faa641a15dfc9940d2a2add52d)
|
|
|
|
| |
deletes
|
| |
|
|
|
|
| |
(cherry picked from commit 5df159a364ec3a94d1e1ae01c70e3ca33cb10b4f)
|
|
|
|
|
|
|
|
| |
Allows multiple writes to be performed on the CollectionCatalog while reusing a single copy of the CollectionCatalog instance.
Used when opening the catalog to avoid quadratic behavior in the number of Collections in the catalog.
(cherry picked from commit 67ef8cd1ed3d15cee09de7905f7ba9fedfba176d)
|
|
|
|
| |
(cherry picked from commit 3a48403f63ed65e076e0cd44cccf0f89123f8e86)
|
| |
|
|
|
|
|
|
| |
This fixes an invariant failure in DocumentSourceGroup.
(cherry picked from commit 2f4c2c6bdca301934d32a9f3a1f6fc76946f7b11)
|
|
|
|
|
|
| |
migrations commit
(cherry picked from commit 9b27d9c7582ca544b0c27b119107b57b4e2b8d50)
|
| |
|
|
|
|
| |
(cherry picked from commit 5cf8a293567989fcc970fb21cde4a1af111c8b58)
|
|
|
|
| |
concern on sharded clusters
|
|
|
|
| |
opCtx as interruptible
|
|
|
|
| |
(cherry picked from commit 9e931d09ba9a1269bf11a689e06e56f61dab7e21)
|
|
|
|
| |
(cherry picked from commit 9c229633f98be337d0fe676c2b4b7b1aa5cf4f61)
|
| |
|
|
|
|
|
|
| |
resharding::data_copy::fillBatchForInsert in ReshardingCollectionCloner::doOneBatch
(cherry picked from commit 98a21127ae4957594276a35228a95d8d79d6ad25)
|
|
|
|
| |
(cherry picked from commit dd6bc197f73bf0f40234db2be02efc76560693df)
|
|
|
|
| |
(cherry picked from commit cf80c11bc5308d9b889ed61c1a3eeb821839df56)
|
|
|
|
| |
(cherry picked from commit f094005b0b8a878bc039ddae540b8696f70848f7)
|
|
|
|
| |
miss documents when using hashed sharding
|
|
|
|
| |
measurements
|
|
|
|
|
|
|
|
|
| |
This avoids underlying SharedBuffer to be shared for KeyString for
separate indexes. Allows indexes that need to spill to disk to
independently free memory instead of being pinned by unrelated other
indexes.
(cherry picked from commit 5d1078c60e8f24e027dc72d7ceeac4dab49908b9)
|
|
|
|
| |
on the write codepath
|
|
|
|
| |
(cherry picked from commit b2531ed72eb81c7a9e4951e4aab93c7d190d3023)
|
|
|
|
|
|
| |
critical section"
This reverts commit a83b0c692c886a595b27358fc5eb585547e0297a.
|
| |
|
|
|
|
| |
(cherry picked from commit e8aafb99f494c29e6ed05e3696a9cd2bb6b3ea5b)
|
|
|
|
| |
(cherry picked from commit 45bc7a2c06d788a2fddc8601450d4a1fb8dbf489)
|
| |
|
|
|
|
|
|
| |
with the pinned timestamp
(cherry picked from commit 4635c1f58339cbe754ffd908a16e8d6ebe7aa5a0)
|
|
|
|
| |
(cherry picked from commit 2861a64889384880a2f25c93fa3e7ced80805b0d)
|
|
|
|
|
|
| |
metadata refresh in txn
(cherry picked from commit 02add56a2100bef135281938a0cadaf374279f03)
|
|
|
|
| |
(cherry picked from commit 03364be4054d3d411c55a5b712e3d0733865eecd)
|
|
|
|
|
|
|
|
| |
The ARS (AsyncRequestSender) calls ShardRegistry::getShardNoReload() and
can therefore throw a ShardNotFound exception when the ShardRegistry is
still initializing.
(cherry picked from commit dd41a99e2c177d5bd3a38229d85458117480c5d9)
|
|
|
|
| |
their accumulator counterparts (cherry picked from commit ba83473ade3ee39cd9053ed3b2b3bf3e5f64ba28) (cherry picked from commit 2704db5962489e679e2cc8f3556a0c10482271ad)
|
|
|
|
| |
(cherry picked from commit 0bd0ddfb1d6875c3ce4390d30e0566b107256f29)
|
|
|
|
| |
command
|
|
|
|
|
|
| |
coordinator
(cherry picked from commit 20043571a6f3f081073a14a8655971a479f44461)
|